Term Test #2 Notes

Lesson #10

Connivence Sample

• Composed of members of a pop. that are east to reach

Voluntary Response Sample

• Composed of members of a pop. that select themselves to participate

We limit BIAS through Random Sampling

• Chance

• An example of random sampling is called Simple Random Sampling (SRS)

• Another one is called Stratified Random Sample: Pop. into groups of similar individuals (Strata)

  Example: Dividing university students by their year of study

Large Scale Surveys

• Multi-Stage Sample: Randomly select groups from larger groups so that groups are smaller at each stage

  Example: A large company with many offices, departments and employees

More BIASES

• Underecoverage Bias

  • Some groups of individuals are left out

• Nonresponse Bias

  • Individuals that cannot be contacted/do not respond

• Response Bias

  • Answers are false

Lesson #11

Experiments

• Do something to individuals to observe a response

Individuals

• Subjects of the experiment

Factors

• The explanatory variables (categorical)

Treatment

• Variables that the experimenter controls/modifies

Common Experiment Design

• Randomized Comparative Experiment

  • Comparison of two or more treatments and random assignment into treatments groups

    Example: A university decides to compare the progress of STATS 101 students taught in person and online

    

• Completely Randomized Design

  • All individuals are allocated at random to all treatments

Block Design

• Creating blocks of individuals that are similar in some way that is expected to affect the response

Lurking variable : A variable that is not included in the study but may influence the results, potentially confounding the outcomes of the treatments

Example: A study finds that people who own more books tend to score higher on intelligence tests.

Lurking Variable:

Parental Education Level – More educated parents may buy more books and also encourage intellectual activities, which improve intelligence test scores

Double Blind Experiment

• Neither the individual nor the experiments know the treatment the individual was assigned to

All experimental designs follow three guiding principles:

• Control

• Randomization

• Replication

When analyzing results we look for results that are Statistically Significant

Lesson #12

Probability Model

• All possible outcomes

• The probability of each outcome

Disjoint events

• Two or more events that cannot happen at the same time

  Example: Rolling a Die – Getting a 3 and a 5 on the same roll is impossible

In probability, we have two types of models: Finite and Continuous

• Finite probability model

Lesson #13

When two events aren’t disjoint